Demos/Videos

1. Multi-source-based Approach for Geolocation in ITV video.  (Worked by Neda Mashhadi and Salem Benferhat)

Discription: From the ITV videos, we seek to produce a graphic representation of the wastewater network of the inspected area. However, the manholes visible in these videos are not identified by precise geographical coordinates, which makes their accurate localization difficult, if not impossible. Our goal is to combine these two sources (GIS and ITV) to obtain as complete data as possible on the wastewater network.

2. From GIS to Graphical Representation for Maintaining Connectivity of Wastewater Network Elements.  (Worked by Omar Et-Targuy, Carole Delenne, Salem Benferhet, Ahlame Begdouri, Thanh-Nghi Do, Thanh Ma, and Salem Benferhat)

Discription: .This research discusses the limitations of the separate databases approach for representing wastewater networks in Shapefile and proposes a novel graph-based approach. In particular, each component of the network, such as manholes, structures, pumps, etc., is represented as a node in the novel graph, while the pipes represent the connections between them. The validation of this approach, using five real datasets, confirms its ability to connect the various components of the wastewater network via a graph-based representation.

3. AQuaNetCT: Detecting and Representing Urban Drainage Networks from Video Analysis. (Worked by Thanh Ma, Minh-Thu Tran-Nguyen, Thanh-Nghi Do, and Salem Benferhat)

Discription: Effective management of urban drainage networks is essential for ensuring resilience and infrastructure reliability. This research presents AquaNetCT, an automated framework designed for the detection, analysis, and visualization of manholes in Can Tho city, Vietnam. The system utilizes computer vision for manhole detection/classification and extracts geospatial and textual metadata through OCR. This framework enhances the efficiency of inspection workflows, contributing to the operational sustainability of smart city drainage management.